Add pod wait function in k8s integration test, for pd driver start and also prepull image daemonset

[jingxu97]

See comments here #613 (comment)

Currently workflow, it waits for certain of of time for prepull calls driver or pr pd csi driver part.

[mattcary]

/assign

I have a PR ready that's waiting on the ADC one (https://github.com/mattcary/gcp-compute-persistent-disk-csi-driver/tree/wait)

It conflicts with the ADC change so it was complicated to send it out for review.

[mattcary]

/open

@jingxu97 do we need to do something for the "prepull image daemonset"? We wait on the daemonset deployment, I'm not sure what you mean buy "prepull"

[jingxu97]

@mattcary prepull script is starting a daemonSet with specific container image. The purpose is to download image by running those containers. We want to do this before running tests so that all the images are already local when tests start to run. Otherwise, it will definitely exceed 5 mins timeout for pods to start which is coded in the tests.

Yes, we should use your wait script to wait for those daemonSet and then delete them

[jingxu97]

somehow it cannot be reopened?

[mattcary]

/reopen

[k8s-ci-robot]

@mattcary: Reopened this issue.

In response to this:

/reopen

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[mattcary]

There we go :-)

I just sent a PR for the daemonset issue